Biochemistry 1984, 23, 3659-3662
3659
Base Pairing in Bacillus subtilis Ribosomal 5s RNA As Measured by Ultraviolet Absorption and Fourier-Transform Infrared Spectrometry? Lee-Hong Chang, Kent 0. Burkey,* James 0. Alben, and Alan G. Marshall*
ABSTRACT:
Ultraviolet (260 and 280 nm) and Fouriertransform infrared (FT-IR) spectra of Bacillus subtilis ribosomal 5s RNA have been acquired between 20 and 90 "C. In the presence of added Mg2+, the average UV melting midpoint, T,, is 60 (A260) or 62 "C (Azao),resolving into two components (T, = 54 and 68 "C). In the presence of 10 mM Mg2+, the normalized A260 increases by about 5%, and the average T , increases to 70 "C (A260 or A2ao), resolving into components at 63 and 73 "C at 260 nm but not resolved at 280 nm. From the difference of the 5s RNA FT-IR spectra between 90 and 30 "C, the number of base pairs in B. subtilis 5s RNA was determined by the procedure outlined in the accompanying paper [Li, S.-J., Burkey, K. O., Luoma, G. A., Alben, J. O., & Marshall, A. G. (1984) Biochemistry (preceding paper in this issue)]. Addition of 10 mM Mg2+ increases the number of A-U pairs by 1 (from 1 1 to 12) and the
number of G-C pairs by 2 (from 15 to 17). FT-IR melting curve midpoints show that addition of Mg2+ increases the melting point for both A-U and G-C pairs in B. subtilis 5s RNA. The A-U pairs melt before G-C pairs (56 vs. 64 "C) in the absence of Mg2+,but both types of pairs melt at the same temperature (67 vs. 70 "C) in the presence of Mg2+. Both UV and FT-IR results thus indicate that Mg2+produces only a slight increase in base stacking and base pairing in B. subtilis 5.9 RNA but Mg2+significantly increases the stability of the less stable base pairs (principally A-U pairs). Among three proposed universal secondary base-pairing schemes for ribosomal 5s RNAs, the Studnicka and Luoma-Marshall models best account for the observed high total number of base pairs, and the latter more closely predicts the high proportion of G-C pairs inferred from the present UV and IR hyperchromism results.
I n the preceding paper (Li et al., 1984b), ultraviolet hyperchromism and Fourier-transform infrared (FT-IR) spectra were analyzed to give estimates for the total number and type (A-U, G-C, and G-U) of base pairs in eukaryotic ribosomal 5s RNA (wheat germ). In this paper, we present a similar analysis for base pairing in a prokaryotic 5s RNA from Bacillus subtilis.
concentration (from was 0.76 mM. At that concentration, the proton FT-NMR spectrum (work in progress) is well resolved, showing no evidence of aggregation.
Materials and Methods 5s RNA. B. subtilis 5s RNA was isolated as described elsewhere (Li et al., 1984a). The basic three-step procedure consisted of phenol/sodium dodecyl sulfate (SDS) extraction, followed by DE-32 ion-exchange chromatography and largescale gel filtration chromatography on Sephadex G-75. Preparation of 5s RNA Samples. 5s RNA samples were prepared as previously reported (Burkey et al., 1983), except that during dialysis, one change of buffer was used and about 50 mg of Chelex (mesh 50-100, Sigma) was added to the dialysis buffer when preparing samples without MgC12. Ultraviolet Spectrometry. UV melting profiles at 260 or 280 nm were obtained on a Beckman DU-8 spectrophotometer with the T, Compuset accessory, in single-beam mode. A260 or A280 was recorded at 1 "C intervals from 25 to 95 "C, with a temperature ramp of 1 "C/min. UV absorbance vs. wavelength at 25 or 85 "C was measured on a wavelength-scan Compuset. Initial ,4260 was 0.72 (-Mg2+) or 0.68 (+Mg2+), with Azso = 0.36 (-Mg2+) or 0.28 (+Mg2+), corresponding to concentrations of ca. 0.03 mg/mL. Fourier- Transform Infrared Spectrometry. All measurements followed the procedures of the preceding paper. Sample From the Department of Chemistry (K.O.B., L.-H.C., and A.G.M.), the Department of Biochemistry (A.G.M.), and the Department of Physiological Chemistry (J.O.A.), The Ohio State University, Columbus, Ohio 4321 0. Received October 25, 1983; revised manuscript received February 7, 1984. This work was supported by U S . Public Health Service grants to A.G.M. (NIH IR01-GM-29274) and J.O.A. (NIH 1 RO1-HL- 17839 and HL-28 144). *Present address: USDA/Agricultural Research Service, Plant Science Research, North Carolina State University, Raleigh, N C 27607.
0006-2960/84/0423-3659$01.50/0
Results and Discussion ,4260 UV Hyperchromism. Figure 1 shows the UV absorbance melting profiles, normalized to room-temperature ,4260 (or Azao), for B. subtilis 5s RNA in the absence and presence of 10 mM Mg2+. The original UV hyperchromism = A2,(90 "C) - A2,(25 "C) increases from 0.19 in the absence of Mg2+ to 0.22 in the presence of 10 mM Mg2+. On the basis of a value of 0.30 for hyperchromism from maximal doublestranded stacking (i.e,, maximal base pairing) and a 10% minimal contribution from single-stranded stacking (Boedtker & Kelling, 1967), an upper-limit base-pair number is (0.22 - 0.022)/0.30 X 100 = 66%of the 116 bases, or about 38 base pairs (Table I) in the presence of Mg2+. In the presence of Mg2+,the A2, melting midpoint (T, = 70 "C) for B. subtilis 5s RNA is 4-5 "C lower than that for tRNAne and about the same as that for wheat germ 5s RNA (see preceding paper). As shown by Ohta et al. (1983), the barely discernible stepwise character of UV melting curves such as Figure 1 can be magnified by plotting dA260/dTvs. T for the same data. For example, the A260melting curve derivative (not shown) for B. subtilis 5s RNA is resolved into two distinct components with T , = 54 and 68 "C (no Mg2+) and T, = 63 and 73 "C (10 mM Mg2+). It is also clear from Figure 1 that Mg2+increases the ,4260 of B. subtilis 5s RNA by about 5% [Le., approximately the same as for yeast 5s RNA (Luoma et al., 1980)l. The corresponding increase for wheat germ 5s RNA is about 1%. Thus, although all three 5s RNAs appear to have the same number of base pairs once Mg2+is present, Mg2+appears to be more necessary for maximal secondary base pairing in B. subtilis and yeast than in wheat germ. A280 UV Hyperchromism. Azso hyperchromism largely reflects unstacking of G-C pairs (Fresco et al., 1963; Van et al., 1977). For B. subtilis 5s RNA, T, increases from 62 to
0 1984 American Chemical Society
3660 B I o c H E M I s T R Y
C H A N O ET AL. I
"
"
"
'
I
0 I
,
240
,
,
,
,
.
260 280 h Inrn1
FIGURE 2 UV hvoerchromism suectrum for R. subfilis 5s RNA (-1
compared to CUI& calculated 'from standard spectra for 50 (;.-j or 60%G-C pairs (--).
Temperature ('C) 1.5.
1.4
1.3
"80
A
Eo
1.2
1.1
1.0
Temperofure ('C) Normalized A, (top) and A,, (bottom) hyperchromism for R. subfilis 5s RNA in the absence ( 0 ) and presence ( 0 )of 10
FIGURE 1:
mM Mgzc. Table I Number of Secondary Base Pairs in 8. subfilis 5s RNAD
source cloverleaf mcdel revised Fox-Woese model Studnicka et al.Nishikawa-Takemura mcdel UV hyperchromism (+hi&
. -Mg2+ ... FT-IQ
A-U 10 6 7 17
G-C 20 17 23 21(GC
G-U
total 4 34 2 25 4 34
+ GU)
38
1
(IO) 26 (36) I1 15 I2 17 (81 . , 29 (37) 'Precision of FT-IR A-U and G-C values is *IO% G-Uestimates are much less precise (see text). +ME,' -
. ,
70 "C and the melting changes from the biphasic to monophasic on addition of lOmM Mgz+. Thus, the effect of Mg2+ appears to be to increase the base stacking and stability of the less stable segments, without much effect on other already highly stacked and stable segments. Similar effects have been observed from differential scanning calorimetry of tRNAVall (Privalov et al., 1975) and Escherichia coli 5s RNA (Matveev et al., 1982). Hyperchromism us. Wauelength-G-C/A-U Ratio. Figure 2 shows the UV hyperchromism spectrum of E . subtilis 5s RNA (solid line), compared to spectra simulated from 5050 and 6040 relative contributions from G-Cand A-U base pairs (Fresco et al., i963; Luoma et al., 1980). As can be seen from the simulations, the best match for the experimental curve is found to be 55 f 5% G-C pairs (Table I). In this estimate, single-stranded stacking, G-U pairs, nearest-neighbor effects, and end-of-helix effects are not considered, FT-IR Spectra Determination of the number of A-U and G-C base pairs in RNA from FT-IR data has produced results inconsistent with other techniques, other laboratories,and other RNA species (see preceding paper). In early experiments with dispersive IR spectrometers, IR spectral data points were sampled at intervals of approximately 4 cm-I in the range 1600-1700 cm-l (Thomas, 1969). Schernau & Ackermann (1977) fitted a linear combination of the spectra of poly(rA).poly(rU), poly(rG)-poly(rC), 5'-3' ApA, 5'-3' UpU, 5'-3' CpC, and 5'-GMP directly to the RNA spectrum in the 157&1670-cm-l region. Deviations from a best fit straight line for the region above 1670 cm-' were attributed to insensitivity of the IR frequency and intensity to the verticalstacking interactions of the base residues, to the length of the helical regions, and to the base sequence. Still, the deduced base-pair number of 56 for E. coli 5 s RNA in the presence of Mgz+ (Appel et al., 1979) was unrealistically high. B6hm et d.(1981) introduced the idea of fitting the RNA difference spectrum (90 minus 20 "C) to the IR hyperchromism spectra for A-U and G-C pairs in 1:l mixtures of complementary homopolymers, on the basis of the intensities of the 1575- and 162o-Cm-l peaks. Burkey et al. (1983) improved the reliability of the base-pair estimate by introducing FT-IR in place of dispersive IR and by basing the calculation on all spectral intensity data within * 5 cm-l of the four largest peaks in the difference spectrum, namely, peaks at 1574, 1618, 1660, and 1688 cm-I. Figure 3 shows the FT-IR extinction coefficient spectra of E. subtilis 5s RNA in the absence of Mg2+at 90 (Figure 3A) and a t 30 "C (Figure 3B) and the difference of these two spectra, also defined as "base-pair spectrum" (Figure 3C). Analogous spectra are shown in Figure 4 for E . subtilis 5s RNA in the presence of 10 mM Mg2+. G-C, A-U, and G-U base pairs were determined by successive fits to the peaks at
VOL. 2 3 , NO. 16, 1984
FT-IR O F B . S U B T I L I S 5s R N A
I
I
B subltlis 5 s RNA
I
3661
B subtilts 5s RNA IOmM MgC12
E. subtilis 5 5 RNA NO MgC12
- 0.25
,
0.25 D Temperature ("C)
Number of A-U (0)and G-C (0) base pairs of B. subtilis 5s RNA as a function of temperature: (A) in the presence of Mg2+; (B) in the absence of Mg2+.
FIGURE 5:
I
. 1400
I 1500 1600 1700 FREQUENCY ( C M - ' )
1800
FT-IR spectra of B. subtilis 5s R N A in the absence of (A) FT-IR spectrum of B. subtilis 5s R N A at 90 OC; (B) FT-IR spectrum of B. subtilis 5s R N A at 30 OC; (C) difference spectrum, or base-pair spectrum, (A) - (B), representing the infrared intensity due to base pairing of the R N A bases; (D) simulation of (C), computed from the reference A-U and G-C base-pair spectra with base-pair numbers listed in Table I (14 A-U and 17 G-C pairs); (E)difference spectrum, (D) - (C), representing the difference between FIGURE 3:
Mg?
experimental and simulated base-pair content. B subtilis 5s RNA +Mg"
075 A
Table 11: Melting MidDoints ("CI of B. subtilis 5 s RNA type of base pair -Mg2+ +Mg2+ UV hyperchromism A-U, G-C, 61 (54 70 (63 G-U and 68) and 73) FT-IR A-U 56 67 G-C
W ,
A
025-
0125 ..
0
.
. ...
.-
--O0.125 2F'====I 1
,
1400
1500
1600
1700
I800
I
FREQUENCY ( C M - ' )
FT-IR spectra of B. subtilis 5s R N A in the presence of ( A X ) and (E) as in Figure 4; (D) simulationof (C), computed from the reference A-U and G-C base-pair spectra with base-pair numbers listed in Table I. FIGURE 4:
Mg?
1505, 1620, and 1575 cm-I (see preceding paper). In this data reduction, the G-C base-pair estimate should be the most precise, since the IR intensity at 1505 cm-' should be affected only by G-C stacking. The G-U estimate is likely to be highly
70
imprecise (*50%), since it is based on previously estimated G-C and A-U base-pair numbers and on the (untested) assumption that G-C and G-U exhibit the same hyperchromism at 1575 cm-I (see preceding paper). Although the absolute numbers of A-U and G-C base pairs obtained by FT-IR and UV differ (Table I), it is encouraging to find that the ratio, G-C/(A-U G-C) = 0.55 from UV hyperchromism, closely matches the analogous ratio of 17/(12 17) = 0.59 from FT-IR (see Table I) for B. subtilis 5 s RNA in the presence of 10 mM Mg2+. Finally, Table I shows that addition of 10 mM Mg2+ increases the number of A-U pairs by one (from 11 to 12) and the number of G-C pairs by two (from 15 to 17), much as for eukaryotic wheat germ 5 s RNA (see preceding paper). FT-IR Melting Profiles. Figure 5 shows the melting of the G-C and A-U base pairs in B. subtilis 5s RNA, determined from FT-IR extinction coefficients at 1505 and 1620 cm-'. The melting midpoint (Table 11) increases from 56 to 67 "C for A-U pairs but only from 64 to 70 "C for G-C pairs on addition of 10 mM Mgz+. Thus, the stabilizing effect of Mg2+ acts principally on A-U pairs, probably because the G-C-rich regions are already highly stable even before addition of Mg2+. The opposite is true for wheat germ 5s RNA, presumably because G-C pairs are more uniformly dispersed in that case. Secondary Structural Models. Figure 6 shows three generalized secondary base-pairing schemes for prokaryotic 5s RNA, each adapted to the B. subtilis 5s RNA base sequence. Model A is the highly base-paired "cloverleaf" model proposed by Luoma & Marshall (1978a,b) and can be adapted to all known prokaryotic and archebacterial and eukaryotic 5 s RNAs as well as to eukaryotic 5.8s RNA. Model B is the original Fox & Woese (1975) model for prokaryotic 5 s RNA, as revised by Noller et al. (1979) for Gram-negative prokaryotic 5s RNA (e.g., E . coli, whereas B. subtilis is a Grampositive prokaryote) and by Luehrsen & Fox (198 1) for eukaryotic cytoplasmic 5 s RNA. Model C is a different variant
+
-025-
64
+
3662
C H A N G ET A L .
BIOCHEMISTRY 5' 3'
Acknowledgments Major components of the interferometer were made available (to J.O.A.) by Dr. Robert L. Berger, Chief, Biophysical Instrumentation Section, Laboratory of Technical Development, NHLBI, National Institutes of Health. Registry No. Magnesium, 7439-95-4.
References
5' 3'
C-OH I16
:E K
I
I
Proposed secondary structuremodels of B. subtilis 5s RNA. Sites of cleavage by TIRNase are shown by small arrows: (A) cloverleaf (Luoma-Marshall) model; (B)revised Fox-Woese model; (C) Studnicka et a1.-Nishikawa-Takemura model. FIGURE 6:
of the Fox-Woese model exhibiting additional base pairing for prokaryotic 5 s RNA (Studnicka et al., 1981) and for eukaryotic 5s RNA (Nishikawa & Takemura, 1978). The arrows inserted onto each model denote points of cleavage by T, RNase (Morell et al., 1967) and probably correspond to single-strandedsegments. Table I lists the basepair numbers predicted by each of the three secondary structural models, for comparison to the values deduced from UV and IR hyperchromism data. Both the cloverleaf and Studnicka models account for the high total number of base pairs, but the cloverleaf model more closely predicts the high proportion of G-C base pairs inferred from the UV and IR hyperchromism experiments. Of course, all three models are intended to account only for secondary base pairs, and there may well be a significant number of tertiary base pairs: 6 of the 27 base pairs in tRNAPheare tertiary pairs (Sussman & Kim, 1976).
Appel, B., Erdmann, V. A., Stulz, J., & Ackermann, Th. (1979) Nucleic Acids Res. 7, 1043-1057. Boedtker, H., & Kelling, D. G. (1967) Biochem. Biophys. Res. Commun. 29, 759-766. Bahm, S . , Fabian, H., Venyaminov, S. Y., Matveev, S. V., Lucius, H., Welfle, H., & Filimonov, V. V. (198 1) FEBS Lett. 132, 357-361. Burkey, K. O., Marshall, A. G., & Alben, J. 0. (1 983) Biochemistry 22, 4223-4229. Erdmann, V. A. (1976) Prog. Nucleic Acid Res. Mol. Biol. 18,45-90. Fox, G. E., & Woese, C. R. (1975) Nature (London) 256, 505-507. Fresco, J., Klotz, L., & Richards, E. G. (1963) Cold Spring Harbor Symp. Quant. Biol. 28, 83-90. Li, S.-J., Chang, L.-H., Chen, S.-M., & Marshall, A. G. (1984a) Anal. Biochem. (in press). Li, S.-J., Burkey, K. O., Luoma, G. A., Alben, J. O., & Marshall, A. G. (1984b) Biochemistry (preceding paper in this issue). Luehrsen, K. R., & Fox, G. E. (1981) Proc. Natl. Acad. Sci. U.S.A. 78, 2150-2154. Luoma, G. A., & Marshall, A. G. (1978a) Proc. Natl. Acad. Sci. U.S.A. 75, 4901-4905. Luoma, G. A., & Marshall, A. G. (1978b) J. Mol. Biol. 125, 95-105. Luoma, G. A., Burns, P. D., Bruce, R. E., & Marshall, A. G. (1980) Biochemistry 19, 5456-5462. Matveev, S. V., Filimonov, V. V., & Privalov, P. L. (1982) Mol. Biol. 16, 990-999. Morell, P., Smith, I., Dubnau, D., & Marmur, J. (1967) Biochemistry 6, 258-265. Nishikawa, K., & Takemura, S . (1974) FEBS Lett. 40, 106-109. Noller, H. F., & Garrett, R. A. (1979) J. Mol. Biol. 132, 621-636. Ohta, S., Maruyama, S., Nitta, K., & Sugai, S. (1983) Nucleic Acids Res. 11, 3363-3373. Privalov, P. L., Filimonov, V. V., Venkstern, T. V., & Bayev, A. A. (1975) J. Mol. Biol. 97, 279-288. Schernau, U., & Ackermann, Th. (1 977) Biopolymers 16, 1735-1 745. Studnicka, G . M., Eiserling, F. A,, & Lake, J. A. (1981) Nucleic Acids Res. 9, 1885-1904. Stulz, J., Ackermann, Th., Appel, B., & Erdmann, V. A. (1981) Nucleic Acids Res. 9, 3851-3861. Sussman, J. L., & Kim, S. H. (1976) Biochem. Biophys. Res. Commun. 68, 89-96. Thomas, G . J., Jr. (1969) Biopolymers 7, 325-334. Van, N. T., Nazar, R.N., & Sitz, T. 0. (1977) Biochemistry 16, 3754-3759.
